Molecular structure and reactivity of Vitamin B6/salicylaldehyde containing model enzymes by Andrew Gilchrist Sykes A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Chemistry Montana State University © Copyright by Andrew Gilchrist Sykes (1984) Abstract: Deuterium exchange of the two glycine protons in sodium bis(pyridoxylideneglycinato)cobaltate(III) is examined. Second order rate constants for exchange in a carbonate/deuterated bicarbonate buffer in D2O are determined, and activation parameters are calculated accordingly. Glycine protons exhibit differing reactivities, the faster proton exchanged roughly ten times the rate of the slow proton over a thirty degree temperature range. The difference in reactivities is attributed to greater ς-π overlap of the fast proton in the transition state, and both NMR of the complex in solution and crystallographic evidence support the different orientations of glycine protons to the neighboring pi system. Activation parameters for the fast proton are ΔH±= 9.9±2 kcal/mole and ΔS±= -28±7 e.u., and ΔH±= 14.5±1 kcal/mole and ΔS±= -17±4 e.u. for the slow proton. These energies differ in numerical magnitude from activation parameters done in a previous study. MOLECULAR STRUCTURE AND REACTIVITY OF VITAMIN Bg/SALICYLALDEHYDE CONTAINING MODEL ENZYMES
by
Andrew Gilchrist Sykes
thesis submitted in partial fulfillment of the requirements for the degree
of
Master of Science
in
Chemistry
MONTANA STATE UNIVERSITY Bozeman, Montana
October 1984 APPROVAL
of this thesis submitted by
Andrew Gilchrist Sykes
This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies.
Date Chairperson, Graduate Committee
Approved for the Major Department
Date Head, Major Department
Approved for the College of Graduate Studies
Date Graduate Dean iii
STATEMENT OF PERMISSION TO USE
In presenting this thesis in partial fulfillment of the requirements for a master's degree at Montana State
University, I agree that the Library shall make it available to borrowers under rules of the Library. Brief quotations from this thesis are allowable without special permission, provided that accurate acknowledgment of source is made.
Permission for extensive quotation from or reproduction of this thesis may be granted by my major professor, or in his absence, by the Director of
Libraries when, in the opinion of either, the proposed use of the material is for scholarly purposes. Any copying or use of the material in this thesis for financial gain shall not be allowed without my written permission.
Signature
Date To my grandparents with love.
Edith Reed VanHorn and Bert Allison VanHorn
Dorthy O'Neil Sykes and Edwin Gilchrist Sykes V
ACKNOWLEDGMENTS
I would like to take this opportunity to thank the
following people for their support and invaluable
assistance without which this project would have not been possible.
To Ray Larson, I extend my gratitude for his genius
in crystallography. I also wish to thank Jim Fischer of
Western New Mexico College for his kinetic and synthetic
knowhow and for getting the ball rolling in the first
place. My colleagues in research, Mark Anderson, Scott
Busse, and Eric Peterson deserve.great praise as well for
an enjoyable .and worthwhile two years.
I would especially like to thank Dr. Edwin H. Abbott
for his never ending positiveness and direction, combined with good philosophy and patience in guiding me through
this project.
Finally, I express my deep gratitude for my parents,
Richard and Virginia Sykes, for their continual love and
encouragement. vi
TABLE OF CONTENTS
Page
LIST OF TABLES...... vii
LIST OF FIGURES...... viii
ABSTRACT...... %
INTRODUCTION...... I
STATEMENT OF OBJECTIVES...... 12
KINETIC RESULTS...... 13
CRYSTALLOGRAPHIC RESULTS...... 25
DISCUSSION...... 35
EXPERIMENTAL...... 47
NOTES AND REFERENCES...... 54 vii
LIST OF TABLES
Page
Table I. First order rate data, A proton, 13°C...... 20
Table 2. First order rate data, A proton, 20°C...... 20
Table 3. First order rate data, A proton, 32°C...... 21
Table 4. First order rate data, A proton, 42*C...... 21
Table 5. First order rate data, B proton, 3*C...... 22
Table 6. First order rate data, B proton, 13 C...... 22
Table 7. First order rate data, B proton, 21 C...... 23
Table 8. First order rate data, B proton, 33°C..... 23
Table 9. Second order rate data...... 24
Table 10. Bond angles (e)...... 31 Table 11. Angles in the coordination polyhedron around Co 31 Table 12. Bond lengths (X)...... 32 Table 13. Anisotropic thermal parameters...... 32
Table 14. Atomic coordinates and isotropic thermal parameters...... 33
Table 15. H-atom coordinates and isotopic thermal parameters...... 34
Table 16. Hydrogen bonds...... 34
Table 17. Crystallographic data...... 53 viii.
LIST OF FIGURES
Page
Figure I. Non-enzymatic pyridoxal-Schiff base model.... 2
Figure 2. The Snell Mechanism...... 3
Figure 3. 313-acetoxycholestan-7-one...... 6
Figure 4. Favored paths of enolization and protonization 7
Figure 5. (I) Co(III) pyridoxylidene glycine; (2) substituted Co(III) salicylidene glycine..... • 9
Figure 6. 1H NMR decay of the glycine AB pattern showing deuterium exchange of the fast B proton...... 14
Figure 7. NMR decay of the glycine AB pattern showing deuterium exchange of the slow A proton...... 15
Figure 8. Typical first order rate plot of integrated NMR peak areas vs. time...... 16 9. Second order kinetic plot of the slow A proton Figure 17
Figure 10. Temperature dependence for the second order rate constants for exchange of the glycine protons of Co(III) pyridoxylidene glycine. Both protons have correlations greater than 19
Figure 11. Structure of the complex anion with averaged bond lengths and angles...... 25
Figure 12. Projections of ligands on CoO (1)0 (2) Nd) and CoO (I') O (2') N d 1) planes...... 27
Figure 13. Crystal packing of TMAtCo(PGly),) viewed along the b axis. TMA cations are shaded dark, and water molecules are labeled and connected by 30
Figure 14. Co(III) pyridoxylidene glycine illustrating axial and equatorial positions of the 36 I
ix
LIST OF FIGURES-Continued
Page
Figure 15. Similiar puckering of 5-membered metalocycle of Co(III) 3-methyl salicylidene threonine... 36
Figure 16. Distortion of one ligand of Co(III) salicylidene glycine while the other remains 38 virtually planar...... Figure 17. Distorted A isomer of Co(III) pyridoxylidene 38 valine...... Figure 18. Allylic fragment of Co (PGly)2 demonstrating the dihedral angle dependence...... 40
Figure 19. AB pattern of the glycine protons along with long distant, allylic coupling by the azomethine proton. Other peaks are the S-Cf^ group and a suppressed solvent peak...... 42
Figure 20. 1H NMR spectrum of Co(III) pyridoxylidene 49 glycine...... X
ABSTRACT
Deuterium exchange of the two glycine protons in sodium bis(pyridoxylideneglycinato)cobaltate(III) is examined. Second order rate constants for exchange in a carbonate/deuterated bicarbonate buffer in D^O are determined, and activation parameters are calculated accordingly. Glycine protons exhibit differing reactivities, the faster proton exchanged roughly ten times the rate of the slow proton over a thirty degree temperature range. The difference in reactivities is attributed to greater o —x overlap of the fast proton in the transition state, and both NMR of the complex in solution and crystallographic evidence support the , different orientations of glycine protons to the neighboring pi system. Activation parameters for the fast proton are Afftr= 9.9+2 kcal/mole and AS = —28+7 e.u., and AH+ = 14.5±1 kcal/mole and AS*= -17+4 e.u. for the slow proton. These energies differ in numerical magnitude from activation parameters done in a previous study. I
INTRODUCTION
Involved in the metabolism of amino acids is the
cleavage of some chemical grouping to the a-carbon.
Cleavage of the C^-COOH of Cft-H bond is thermodynamically most unfavourable since pKa values of an unmodified amino
acid alpha carbon lie in excess of 30, rendering AG
values approximately equal to 38 kcal/mole or greater.
Only through some persuasive form of catalysis, can the metabolism of amino acids be realized.
The original discovery leading to the interest in
pyridoxal dependent biological reactions, the key to the
catalysis puzzle, was made in 1934 by Paul Gyorgy of I Western Reserve University. This nutritional factor,
which Gyorgy called vitamin Bg, was subsequently
identified as pyridoxine, one of a number of closely
related compounds in the grouping, pyridoxal among
them. Pyridoxal's utility as a catalytic agent became
apparent when it was recognized formation of a Schiff
base between the cofactor and an amino acid greatly
affected rates of alpha carbon cleavages. In 1952,
Metzler and Snell expanded the field even further when
they published a pyridoxal-dependent, non-enzymatic 2 2 transamination in the presence of metal ions. Numerous enzymatic reactions have now been reproduced using the metal-amino acid-pyridoxal model system proposed by
Snell, all having the basic features shown in Figure I.
O Ii
Figure I. Non-enzymatic pyridoxal-Schiff base model.
A number of these non-enzymatic reactions, paralleling the enzymatic, biological reactions, are condensed in Figure 2, demonstrating the enormous versatility of the model system. Common in all pathways, however, is the initial loss of an electropositive substituent from the alpha carbon. Loss of the substituent and consequent formation of negative charge on the alpha carbon is relieved through two structural features of the model system. One, the conjugated pi system transfers electrons through the molecule and coordinates a proton on the pyridine nitrogen. Two, the metal ion itself is electronegative, reducing electronic 3
R-C-COO" NH2 THE SNELL MECHANISM + M** Metzler. Ikawa. 4 Snell. JACS. Zfi. 648II9S4I.
CHO
DECARBOXYLATION Nf vCH3 REVERSE ALDOL ADDITION
R-CH-"* 'OOC-CH-"* hc\Vm ** HC'Vy** + RCno HOCH2y A y 0
CH3
H + H* k W* I
R-C 1 CO2H 'OOC-C ^-R'
Less el COOH * HCiC "XXC H 3 HOCH2Vf "OCH2Vr y u IL CH3 ^ CH3 H
8 - ELIMINATION B - T - DESATURATION
Figure 2. The Snell Mechanism." 4 charge on the alpha carbon. Both these effects have a catalytic value, lowering the carbanionic character of the alpha carbon and increasing rates of reactions, reactions unlikely without such assistance. Furthermore, the metal ion locks the molecule into a particular constrained conformation as will be discussed further on.
Though metal ions in model systems are not obligatory for amino acid transformations (some do proceed without metal ions present), their presence has generally been found to greatly enhance reaction rates and product formation. In pyridoxal-dependent enzymes themselves, however, it is unlikely metal ions play an active role. Certainly it is the prosthetic groups of pyridoxal combined that leads to the cofactors utility.
The pyridinium nitrogen, the 5-hydroxy or phosphate group, plus the carboxylic group of the amino acid all must fit or bind into the particular pattern of the enzyme in such a way as to lower the pKa value of the relevant carbon acid-whether or not in the presence of metals.
Even despite limiting evidence of metalloenzymes, the remarkable work commenced by Snell and his associates
in the early 501s has led to a group of metal complexes
successful in reproducing the reactions of metal-free or metal-containing biological molecules. The scope 5 delimiting the types and methods of non-enzymatic model reactions is now rather complete. What remains to be done is significant research activity in coordination chemistry and catalysis, and even more remains to be understood about the structural and electronic features of the complexes themselves. To provide a clearer understanding of reactivities, precise structural determinations need to be made along with further detailed kinetic studies of the various transformations as well. Helping understand the factors controlling reactivities of substituents at amino alpha carbons in complexes such as in Figure I, then, is the general intent of this thesis.
The hypothesis of Snell concerning the mechanism of amino acid transformations emphasizes the function of the cofactor and metal ion in weakening the sigma bonds around the alpha carbon, but it does not address the question about which bond is cleaved most easily. For example, why the preferential loss of one alpha carbon substituent over another? In the late sixties, H.C.
Dunathan attributed the stereochemical relationship of the sigma bond to the adjacent pi system as being responsible for determining reaction specificity.^ Since loss of a group from the amino acid alpha carbon results
in carbanibn formation and the extension of the pyridoxal 6 imine system, an increase in delocalization energy occurs. If this gain in delocalization energy is to aid in bond breaking, a geometry placing the bond to be broken perpendicular to the pyridoxal imine system
(coplanar with the pi system) is most highly favoured.
Thus, transition states where sigma bond to pi system interactions are as coplanar as possible will enhance reactivities. It is expected bond breakage assisted by delocalization energy will have lower enthalpies of activation than sigma bonded substituents bonded more perpendicular to the pi system.
Figure 3. 3/3-acetoxycholestan-7-one.
Early work done by Corey on the stereochemistry of enolization of 3/9-acetoxycholestan-7-one (Figure 3) showed clear preference for bromination to occur axially at the 6-methylenic carbon.5 Opposing this effect is the classical steric argument, more I,3-interactions, which directs a substituent as large as bromine towards the less crowded equatorial position. Obviously, the axial 7 product is formed kinetically rather than for thermodynamic reasons, and it is thought in the transition state, since the enolization-ketonization process is stabilized by bonding between alpha and carbonyl carbons involving ct- tt delocalization, there should be a demonstrated preference for loss or gain of an axial alpha substituent. Subsequently an axial hydrogen is lost in enol formation and return of bromine occurs likewise in the more hindered axial position. This is the same phenomenon as described by Dunathan above.
Figure 4. Favored paths of enolization and protonization of testosterone. 8
Malhotra and Ringold studied the kinetically controlled enolization of testosterone (Figure 4 ) Both
NaOD and CH3COOD in D2O catalyzed deuterium exchange at the 2-position immediately adjacent to the carbonyl group, and exchange, in both cases, occurred axial as well. Deuterium chloride catalysis resulted in the preferential loss and exchange of the C-6 proton. This change in specificity is attributed to base strength.
Using strong base, CHgCOO or OD- , the determining factor in the transition state is the acidity of the methylenic protons. C-2 protons next to the carbonyl being are more acidic than the C-6 protons, and at the C-2 carbon, the axial proton is more labile than the equatorial proton for reasons put forth above. With DCl catalysis however, since DgO is the strongest base present, considerable C-H bond stretching in the transition state will lead to a greater resemblance to enol. Acidity (C-2 vs. C-6) will assume little relative importance, and stability now hinges on the respective enols themselves. It should be noted though that even with DCl catalysis, the more acidic axial C-6 proton was lost and preferentially replaced with retention of configuration.
Recent studies have focused on the kinetics of deuterium exchange between glycine protons of the alpha carbon in the following model systems (Figure 5). 9
R = H7CH3
Figure 5. (I) Co(III) pyridoxylidene glycine; (2) substituted Co (III) salicylidene glycine.
Belokon of the Soviet Union, working with the salicy!aldehydes (2), discovered deuterium exchange of simple ortho-hydroxybenzaldehyde complexed in the previous manner to glycine and cobalt led to monophasic kinetics.7 This is consistent with the crystal structure for bis(salicylideneglycinato)cobaltate(III) (2, R=H), showing two planar ligands. 8 With planar ligands, it is expected the glycine protons having tetrahedral arrangements around carbon would exemplify identical reactivities. The tetrahedral arrangement allows both alpha protons the same orientation with respect to the neighboring pi system, providing the five-membered metalocycle is flat.
The introduction of a methyl group at the 3 position of the benzene ring (2, R=CH3) causes drastic changes in the structure of the molecule. Pyridoxal has the structurally equivalent methyl group on the pyridinium 10 ring but in the 2 position since numbering now begins with the pyridine nitrogen rather than at the aldehyde.
Belokon observes biphasic kinetics for the 3-methyl substituted salicy!aldehyde, the fast proton exchanging more than ten times the rate of the slow proton. Rate constants for the two exchange reactions at 25°C are 5.47
M ^s ^ for the fast and 0.48 M-1S-1 for the slow. The monophasic rate constant is 6.26 (.3) M ^s
Fischer and Abbott in 1978 directed their attention towards how enzymes brought about enhancement of rates, completing a preliminary study of deuterium exchange of
Q the b i s (pyridoxylideneglycinato)cobaltate(III) ion (I).
They presumed the glycine protons around the alpha carbon were forced into different conformations with respect to the azomethine pi system and would consequently exhibit differing reactivities; thereby testing the Dunathan hypothesis. The result of their kinetic work is quite interesting. Indeed, biphasic kinetics were observed. The difference in temperature dependence for the two protons revealed a relatively low AH* and a relatively negative
AS* for the fast proton. For the fast reaction; AH* =5±3 kcal/mole and AS* =-50+15 e.u., and AH =16+2kcal/mole and AS+ =-19±5 e.u. for the slbw reaction. These results are consistent with the Dunathan hypothesis. For the fast reaction, overlap with the azomethine pi system is 11 * possible and A H is considerably reduced. The large negative AS^ requires the arrival of base to be colinear with the pi system and imposes sizable limitations on exactly how the base arrives and abstracts the fast proton.
There are limitations however. Second order rate constants could not be determined because of the experimental conditions, and only pseudo first order constants were used in calculating activation parameters.
Concentrations of reacting complex were very high in order that NMR spectra could be obtained, but this initiated side processes such as base catalyzed proton exchange of one complex by the basic pyridine nitrogen of another. How meaningful the numerical magnitudes of the activation parameters remain in doubt.
No crystal structure has been done with glycine in the 3-methyl salicylaldehyde compound (2, R=CH3), nor with the pyridoxal-glycine complex (I) either. Crystal determinations of complexes using amino acids other than glycine have been completed and exhibit general distortions of the 5-membered metalocycles. Distortion of the 5-membered ring is deemed necessary to relieve ring strain and create the differing orientations of alpha,
carbon substituents that are responsible for reaction
specificities. 12
STATEMENT OF OBJECTIVES
A critical investigation of the kinetics of sodium bis (pyridoxylideneglycinato)cobaltate (III) (NafCo (PGly^ J) was undertaken to improve and ultimately test the findings of Fischer and Abbott. Second order rate constants were obtained and used to calculate activation parameters for the two exchanging alpha protons of the amino acid. At the same time, structural information about the complex was collected by x-ray diffractometry on a crystal fraction of the tetramethy!ammonium salt (TMA{Co (PGly)2 ^ • Sodium salt crystals were unsuitable for x-ray diffraction, necessitating inclusion of the larger TMA cation for crystal growth. It is not expected change in the crystal cation will alter structure of the complex anion. 13
KINETIC RESULTS
Sodium(pyridoxylideneglyclnato)cobaltate(III) exchanges both alpha protons of the amino acid moiety in
D2O under the action of a CO3sZDCO3^ buffer. The rate of I the process is determined from the decay of the H NMR signal of the alpha protons (Figures 6 and 7). Excellent biphasic kinetics are observed, and the slow exchanging proton has been labeled the A proton due to its downfield position in the AB pattern of the NMR spectrum.
Plots of time vs. the logarithm of integrated NMR peak areas followed linear relationships, producing pseudo first order rate constants for both alpha protons.
An exemplary first order rate plot is given in Figure 8.
Tables I through 8 list first order rate constants for both alpha protons at varying temperatures over an approximate 30°C range. Carbonate concentrations plotted against previously obtained first order rate data resulted in second order rate constants (Figure 9). At zero carbonate concentration, intercepts approach zero activity in all cases ruling out base catalysis by water or by the pyridine nitrogen of another complex. However, at pD's higher than 11, there is a deuteroxide dependence as shown by the upward curvature at high carbonate 14
Fast Proton Exchange
9.19 pH 293° K
Figure 6. 1H NMR decay of the glycine AB pattern showing deuterium exchange of the fast B proton. 15
Slow Proton Exchange
(same conditions)
Figure 7 1H NMR decay of the glyicine AB pattern showing deuterium exchange of the slow A proton. 16
concentrations in Figure 9. This is expected since at pD=ll, the ratio of deuteroxide to carbonate concentrations is approximately 1:4. At lower pD's this ratio decreases and the probability of complex encountering a deuteroxide ion is at most seven percent.
Deuteroxide dependence is avoided by keeping pD's less than 11. The highest is 10.61 with most falling between 9 and 10 or lower. Second order rate constants are tabulated in Table 9.
Slope = 4.06 x 10 = k Correlation = 0.998
1,000 2,000 3,000 4,000 Time (sec)
Figure 8. Typical first order rate plot of integrated NMR peak areas vs. time. 17
10.0
[CO=] x 10 (M)
Figure 9. Second order kinetic plot of the slow A proton at 320C.
Measurement of these rates over a suitable temperature difference results in activation profiles as shown in Figure 10. Activation parameters were obtained using the following equation:
k = e m ,RTLeAS4ZRe-AH^RTcI-m i) Nh
k = second order rate constants m = molecularity = 2 R = 1.987 cal/mole N = Avagodro1s # h = Planck's constant in cal.sec T = °K 18
Equation I) reduces to:
% In ( RT I - 2 vs i/f 2) ' Nh ' 4= slope = -Ah /R £ intercept = AS/R
Frequently equation I) is misinterpreted either leaving off the correcting term for concentration units or avoiding em. A good review and explanation of applied activated complex theory can be found in J. Chem Ed., 55,
509(1978), detailing units, rate constants, and molecularity.
Activation errors are derived from the following equations:
T7T 5 = 2R — a 3) I — I
5 is the error in activation enthalpy while a is fractional error in rate constants. Assuming ot= 10% for
the slow reaction (the error in the integration routine),
15% for the fast reaction (error in integration plus the
inability to correctly predict the inflection point
between the A and B peaks of the AB pattern), and a range
in temperature of 30°C:
5 = +1 kcal/mole slow proton
5 = ±2 kcal/mole fast proton 19
Error in Entropy is:
Using 5's above and a 30°C separation in temperature:
o- = ±4e. u. slow proton
cr = +7e. u fast proton
Activation parameters for the two exchanging protons after plotting equation 2) are: AH = 14.5±1 kcal/mole and
AS*= -17±4 e.u. for the slow proton, and AH*= 9.9±2 kcal/mole and AS+= -28t7 e.u. for the fast proton.
.0031 .0032 .0033 .0034 0035 0036
Figure 10. Temperature dependence for the second order rate constants for exchange of the glycine protons of C o (III) pyridoxylidene glycine. Both protons have correlations greater than 0.990. 20 O Table I. First Order Rate Data, A proton. 13 C.
Kinetic PD {CO = } k Cs* ) Temperature Run (corr.)
I 10.42 0.002580 0.000168 11.5 (.988) 2 10.29 0.002050 0.000206 13.9 (.997) 3 10.11 0.001450 0.000122 13.5 (.998) 4 9.94 0.001030 0.000083 14.2 (.976) 5 9.87 0.000891 0.000059 12.5 (.994) 6 9.72 0.000648 0.000060 13.5 (.999) 7 9.55 0.000447 0.000032 11.5 (.977)
Table 2. First Order Rate Data, A proton. 20°C
Kinetic pD {co p k (£•*•) Temperature Run (corr.)
I 10.61 0.003960 0.000461 19.4 (.983) 2 10.20 0.002030 0.000230 21.8 (.980) 3 10.00 0.001390 0.000138 18.7 (.993) 4 9.85 0.001020 0.000117 19.7 (.988) 5 9.84 0.001000 0.000097 21.4 (.988) 6 9.70 0.000745 0.000064 (.991) 7 9.51 0.000495 0.000067 (.977) 8 9.45 0.000434 0.000052 19.7 (.989) 9 9.26 0.000284 0.000027 (.976) 10 9.20 0.000249 0.000041 (.980) 11 8.97 0.000148 0.000017 (.940) 12 8.59 0.000062 0.000011 (.945) 21
Table 3. First Order ,Rate Data, A proton. 3 2 °C.
Kinetic PD . {cof} k (s~^) Temperature Run (corr.)
I 10.45 0.003800 0.001290 - (.975) 2 10.22 0.002650 0.000909 33.8 (.994) 3 9.94 0.001590 0.000510 32.1 (.995) 4 9.87 0.001390 0.000393 3 0.3 (.990) 5 9.85 0.001330 0.000429 33.1 (.991) 6 9.67 0.000920 0.000257 32.4 (.996) 7 9.20 0.000333 0.000134 — (.990) 8 9.13 0.000285 0.000124 31.7 (.970) 9 9.06 0.000243 0.000071 31.7 (.970) 10 9.05 0.000238 0.000097 (.994) ■
Table 4. First Order Rate Data, A proton. 42 “c.
Kinetic PD {CO=} . k (s-1) Temperature Run (corr.)
I 10.38 0.003810 0.002960 41.0 (.992) 2 10.04 0.002200 0.001710 41.5 (.992) 3 9.92 0.001760 0.001410 40.6 (.998) 4 9.89 0.001660 0.001540 41.5 (.990) 5 9.63 0.000990 0.000662 43.0 (.983) 6 9.53 0.000800 0.000532 42.0 (.989) . 7 9.46 . 0.000690 0.000474 43.0 (.987) 8 9.43 . 0.000648 . 0.000377 40.6 (.992) 9 9.37 0.000568 .. 0.000338 43.0 (.993) . 22
Table 5. First Order Rate Data, B proton. 3 C.
Kinetic PD (CO3=J k (s1 ) Temperature Run (corr.)
I 10.08 0.00107 0.000549 4.3 (.995) 2 9.90 0.00073 0.000412 3.2 (.994) 3 9.90 0.00073 0.000384 3.9 (.970) 4 9.77 0.00056 0.000288 3.6 (.990) ■ 5 9.64 0.00042 0.000197 2.9 (.990)
Table 6. First Order Rate Data, B proton. 13 C.
Kinetic PD {cop kts1 ) Temperature Run (corr.)
I 9.82 0.000802 0.000838 12.8 (.995) 2 9.80 0.000769 0.000553 (.988) 3 9.59 0.000488 0.000485 12.5 (.995) 4. 9.40 0.000321 0.000325 12.8 (.993) 5 9.09 0.000160 0.000162 (.984) 6 9.08 0.000156 0.000190 12.8 (.990) 7 9.07 0.000153 0.000174 12.8 (.997) 8 9.06 0.000149 0.000118 (.970) 9 8.92 0.000109 0.000094 (.997) 23 O Table 7. First Order Rate Data, B proton. 21 C.
Kinetic PD
I 9.51 0.000495 0.000756 - (.996) 2 9.42 0.000406 0.000603 20.7 (.985) 3 9.39 0.000380 0.000593 20.7 (.998) 4 9.20 0.000249 0.000406 (.998) 5 8.96 0.000145 0.000261 (.994) 6 8.94 0.000138 0.000164 20.4 (.987) 7 8.93 0.000135 0.000209 (.998) 8 8.66 0.000073 0.000124 (.995) 9 8.59 0.000062 0.000093 (.995)
Table 8. First Order Rate Data, B proton. 33 °C.
Kinetic PD . {cop k(s-1) Temperature Run (corr.)
I 9.30 0.000415 0.001430 31.4 (.986) 2 9.20 0.000333 0.000883 (.981) 3 9.19 0.000325 0.001080 33.8 (.983) 4 9.05 0.000238 0.000801 (.985) 5 9.00 0.000213 0.000598 (.961) 6 8.94 0.000186 0.000545 (.976) 7 8.90 0.000170 0.000638 33.1 (.985) 8 8.89 0.000166 0.000500 (.984) 9 8.83 0.000145 0.000431 (.990) 24
Table 9. Second Order Rate Data.
A Proton -I -I Temperature k (I«m s ) Correlation
12.9 0.076 0.927 20.1 0.114 0.989 32.2. 0.336 0.996 41.8 0.823 0.992
B Proton
3.3 0.54 . 0.989 12.8 0.87 0.962 20.6 1.52 0.995 32.8 3.35 0.960 25
CRYSTALLOGRAPHIC RESULTS
The X-ray determined structure is composed of A and A. bis (pyridoxylideneglycinato)cobaltate(III) anions, tetramethylammoniurn cations, and solvating water molecules. Bond lengths and angles in the two pyridoxylideneglycinate ligands are given in Tables 10,
11, and 12. The anion has an approximate non-crystallographic C2 symmetry. Averaged bond lengths and angles for this structure are given in Figure 11.
I H.. V I 17.‘) 11H. 7
118.7 127.'. I I 1.0
I 18.V
Figure 11. Structure of the complex anion with averaged bond lengths and angles. 26
The distortion of the coordination octahedron of the
Co atom is not large. The octahedron consists of four 0 atoms and two N atoms of the two terdentate pyridoxylideneglycinate ligands. The mean value of the
N(2)CoO(2) angle in the 5-membered metalocycies is
85.9(6), and of the N(2)CoO(I) angle in the 6-membered cycle is 91.8(6). The trans O (I)CoO(2) angle, compared with the ideal angle of 180°, is 177.7(5). The averaged values for bond lengths Co-O(I)=1.869(14),
Co-O(2)=1.917(15), and Co-N(2)=1.856(16)A. These averaged bond lengths and angles are very similiar to those of two x-ray studies done on octahedral Schiff base amino acid 8 10 complexes of Co(III). ' However, these studies exhibited slightly greater distortions from a true octahedral arrangement.
The geometry of the anion is characterized by considerable non-planarity of the pyridoxylideneglycinate ligands. It is convenient to describe the ligand conformations with respect to the dihedral angle between the planes CoO(I)0(2)N(2) and C (2)C (7)N(2)C (9) for both unprimed and primed ligands. The coordination environment around Co is ideally planar and. deviations from planarity in C(2)C(7)N(2)C(9) planes are equally as small. The mean o dihedral angle for these planes in both ligands is 24.5 ,
Projections from the ligand coordination plane (Figure 27
- 0 .4 9
CI2I ) ~ 0.68
- 0.85
- 2.03 - 2.79
- 1.66 -0j7
0.00
CIIO'I
0.00
Figure 12. Projections of ligands on CoO(I)O(2)N(I) and Co0(l')0(2')N(l') planes. 28
12) clearly show ligand distortions. Analogous conformations with ligands tilted away in one direction are present in bis(pyridoxylidenevalinato)nickel(II)^ 7 bis(3-methy1-salicylidenethreonato)cobaltate(III),
and bis(pyridoxyIidenevalinato)manganese (II).11 8 Only bis(salicylideneglycinato)cobaltate(III),
without the 3-methyl group, shows no distortion of the ligands.
The 5-membered metalocycles show no difference in their distortion from planarity: C (9) and C(IO) atoms of both ligands show an average dislocation of 0.50(4) and
0.34(4)A from their CoN(2)0(1)0(2) and CoN(2 1) 0 (11)0(21) planes. H(9a) and H(9a') atoms are axial in orientation to the ligand pi system; corresponding displacements from the ligand coordination plane are 1.50(4) and 1.59(4)A respectively, while H (9b) and H (9b') atoms lie equatorially with displacements of -0.15(4) and 0.00(4)A.
Figure 13 shows a projection of the crystal packing along the b-axis of the unit cell. Numerous hydrogen bonds, involving nitrogens of the pyridine rings, several water molecules, and the phenolic and carboxy-oxygens, help hold the crystal together. Hydrogen-bond lengths are reported in Table 16. Complex anions are hydrogen-bonded within the layers as H (04') points away from the molecule towards the phenolic oxygen, 0 (1), of a differing 29 asymetric unit, while H (04) remains shifted towards the body of the molecule and does not participate in
H-bonding. Surrounded by water molecules, tetramethy!ammonium ions lie in a channel through the crystal and are responcible for a small degree of disorder in the structure as the cations rotate around the central nitrogens. The majority of the disorder is due to wayward water molecules. According to the 0...0 distances, there are eleven H-bonds involving a total of three water molecules split between five positions. O(Wl) has a complete occupancy, but 0(W2), 0(W3), 0(W4), and
0(W5) show partial occupancies of approximately one-half.
Extensive H-bonding between water molecules occurs with each water molecule potentially occupying a variety of positions (i.e. 0 (W4) has four possible H-bonds but a large disorder since when filled with a water molecule only a maximum of two H-bonds will be complete at one time). On top of this are further H-bonds to the basic pyridine nitrogens and both carboxy-oxygens as well.
Anisotropic thermal parameters have been assigned to the cobalt, coordinated oxygens and nitrogens, water oxygens, and atoms of the tetrame thyIammohium cation and are listed in Table 13. Positions and isotopic thermal values for atoms including hydrogens are given in Tables
14 and 15. 30
Oz Figure 13. Crystal packing of TMAtCo(PGly)2> viewed along the b axis. TMA cations are shaded dark, and water molecules are labeled and connected by dashed lines. 31 Table 10. Bond angles (°).
Angle CV C V 1 Ave.
Co-O(I)-C(2) 122.1 (10) 120.6(10) 121.4 Co-O(2)-C(10) 110.3(11) 111.7(13) 111.0 Co-N(2)- C (7) 128.3(11) 126.8(11) 127.5 Co-N(2)- C (9) 112.7(9) 109.4(11) 111.0 N(I)-C(I)-C(2) 123.9(15) 119.8(15) 121.9 N d ) -C (I)-C (6) 114.4(14) 116.1(15) 115.3 Nd) -C (5) -C (4) 122.7(15) 120.9(17) 121.8 N (2) -C (7) -C (3) 120.1(14) 123.5(16) 121.8 N(2)-C(9)- C (10) 104.4(15) 107.6(13) 106.0 0 (I) -C (2) -C (I) 118.5(14) 116.1(14) 117.3 0 (I) -C (2) -C (3) 122.3(13) 123.0(14) 122.7 0(2)-C (10)-0(3) 119.9(17) 124.3(18) 122.1 0(2)-C(IO)-C(9) 122.4(17) 118.9(17) 120.6 0(3)-C(IO)-C(9) 116.9(17) 116.5(16) 116.7 0(4)-C(S)-C(4) 104.8(13) 108.9(14) 106.9 C(D-Nd)-C(S) 117.9(14) 119.9(15) 118.9 C (I) -C (2) -C (3) 119.1(13) 120.9(16) 120.0 C (2) -C (I) -C (6) 121.6(13) 124.1(16) 122.9 C (2) -C (3) -C (4) 119.7(14) 117.7(14) 118.7 C (2) -C (3) -C (7) 123.2(13) 121.7(15) 122.4 C (3)-C(4)'-C(S) 116.4(16) 120.7(16) 118.6 C (3) -C (4)-C(8) 125.6(15) 121.3(15) 123.5 C (4) -C (3) -C (7) 116.8(14) 120.5(15) 118.7 C (5) -C (4) -C (8) 118.0(14) 117.7(17) 117.9 C (7) -N(2)- C (9) 118.5(13) 123.6(14) 121.0 C (11) -N(3)-C (12) 111.2(25) - - C(Il)-N(S)-C (13) 113.1(29) - - C(Il)-N(3)-C(14) 104.9(27) - - C (12)-N(3)-C (13) 112.6(23) - C (12)-N(3)-C (14) 106.3(27) -- C (13) -N(3) -^C (14) 108.1(30) - -
Table 11. Angles in the coordination polyhedron of Co atom.
Angle CV Angle CV
0 (1)CoO(2) 177.0(5) 0(1')CoO(2') 178.4(5) O(I)CoOd') 89.0(5). 0(1')CoO(2) 90.4(5) 0 (1)CoO(21) 91.0(6) 0(1')CoN(2) 91.1(6) 0 (1)CoN(2) 91.6(6) O(I1)CoN(2') 92.0(6) 0 (I)CoN(21) 92.1(6) 0 (2)CoO(21) 89.6(6) 0 (2)CoN(2) 85.5(6) 0 (2)CoN(2') 90.9(6) 0(2') CoN (2). 90.5(6) 0 (2')CoN(2') 86.4(6) N(2)CoN(2') 175.3(6) 32 Table 12. Bond lengths (A). Bond d (A) Bond d (A)
Co-O(I) 1.880(13) C o - O d 1 ) 1.855(14) Co-O(2) 1.920(14) Co-O(2') 1.915(16) Co-N(2) 1.835(15) Co-N(2') 1.877(16) O (I)- C (2) 1.305(17) 0 (1')-C (2') 1.345(21) 0(2)-C(IO) 1.260(25) 0 (2')-c(io') 1.275(22) 0(3) -C (10) 1.318(25) 0(3' )-C(10') 1.269(25) 0(4)- C (8) 1.442(21) 0(4' )-C (8') 1.429(23) N(I)-C(I) 1.370(20) N d ' )-cd') 1.402(24) N(I)-C(S) 1.364(24) N d ' )-C (5') 1.401(24) N(2)-C (7) 1.272(18) N( 2' )-C(7') 1.249(23) N(2)-C(9) 1.506(22) N (2' )-C (9') 1.480(22) C d ) -C (2) 1.387(22) C d ' )- C (2') 1.370(24) C(I)-C(S) 1.483(25) C d ' )-C(6') 1.467(26) C (2)-C (3) 1.404(23) C (2' )-C93') 1.419(23). C (3)- C (4) 1.438(22) C (3 ')-C(4') 1.434(25) C (3)- C (7) 1.441(22) C O ' )-C (7') 1.413(23) C (4) -C (5) 1.455(26) C (4' )-C(5') 1.343(26) C (4)- C (8) 1.507 (26) CU' )-C(8') 1.510(26) C (9)-C(IO) 1.467(26) C (9' )-C(10') 1.498(28) N(3)- C (11) 1.409(42) NO) -C(12) 1.427(36) N(3)-C (13) 1.546(45) NO) -C (14) 1.397(49)
Table 13. Anisotropic thermal parameters (A2XlO3 ). G Atom Ull u22 U 33 U23 H W U12 Co 24(1) 27(1) 38(2) 7(1) -1 (1) 10(1) 0 (1) 24(7) 30(7) 31(6) 6(5) 5(5) 17(5) 0 (2) 40(8) 45(8) 44(7) 15(6) ■ -8(6) 13(6) N( 2) 34(9) 31(8) 18(7) 6(6) 8(7) 3.(7) 0 (1' ) 37(7) 16(6) 41(7) 7(5) 5(6) 8(5) 0 (2' ) 65(10) 36(8) 44(8) -4(6) 5(7) -3(7) N( 2' ) 29(9) 42(9) 18(8) -3(6) , -8(7) -20(7) O(Wl) 107(15) 102(13) 106(13) -18(10) 25(11) 13(11) 0(W2) 110(23) 56(17) 143(25) 17(16) -50(20) 7(16) 0(W3) 192(40) 218(39) 42(19) 0(22) 0(23) 0(34) 0 (W4) 139(34) 133 (29) 134(29) 127 (25) -12(26) -3(25) 0(W5) 131(39) 73(26) 87(29) 37(21) . 37(28) -5(25) NO) 104(17) 81(14) 76(13) -8(11) 8(13) -9(13) C(Il) 255(56) 235(45) 164(36) -71(32) ---55(37) -113(41) C (12) 186(35) 167(29) .65(17) 66(18) -21(20) -38(26) C (13) 309(65) 157(37) 197(40) -5(29) -73(42) 48(38). C (14) 194(46) 415(68) 128(31) 125(37) . 82(31) 121(46) The anisotropic■temperature factor exponent takes the form: -2 7r2(h2a*2Uii +...+2hka*b*U.i2 ) . 33 Table 14. Atomic coordinates (xlO4) and isotopic thermal parameters (A2XlO3).
Atom X y Z U Co 2367(3) 2315(2) 3806(2) 30(1)* O(I) 2559(7) 4775(7) 4775(7) 28(4)* 0 (2) 636(13) 2144(9) 2827(8) 43(5)* 0(3) -577(18) 3001(11) 1725(10) 73(4) 0(4) 1132(13) 6057(8) 5804(8) 39(3) N(I) 5053(18) 4110(11) 7107(10) 50(4) N (2) 2245(15) 3740(9) 3753(8) 27(5) Cd) 4902(20) 3302(12) 6340(11) 32(4) C (2) 4119(18) 3385(11) 5453(10) 25(4) C (3) 3377(19) 4334(11) 5329(11) 26(4) C (4) 3491(23) 5214(14) 6106(12) 48(5) C(S) 4410(21) 5057(14) 6987(12) 42(5) C (6) 5714(23) 2323(13) 6513(13) 47(5) C (7) 2065(19) 4517(12) 4423(10) 26(4) C (8) 2788(20) 6284(13) 6070(12) 37(5) CO) 1395(23) 3961(13) 2828(12) 45(5) C(IO) 522(24) 2965(15) 2433(14) 51(5) 0 (1') 969(12) 2418(7) 4758(7) 31(4)* 0 (2') 3811(16) 2167(9) 2825(8) 49(5)* 0(3') 4781(16) 886(10) 1771(9) 66(4) 0(4') 4065(14) -802(9) 5792(8) 41(3) N d ' ) 610(19) 1616(12) 7071(10) 53(4) N (2 ' ) 2460(16) 839(10) 3750(8) 32(5)* Cd') 568(20) 2220(13) 6319(11) 32(4) C (2') 1116(20) 1808(12) 5456(11) 33(4) C (3' ) 1736(18) 776(11) 5313(10) 23(4) C (4 ' ) 1746(22) 179(14) 6095(12) 46(5) C d ' ) 1228(22) 599(14) 6942(13) 48(5) Cd') -123(24) 3268(14) 6515(14) 58(6) C d ' ) 2251(21) 322(14) 4416(12) 38(5) C (8') 2473(21) -898(13) 6027(13) 43(5) CO') 2978(22) 361(13) 2805(12) 42(5) C(IO') 3883(24) 1203(15) 2425(13) 49(5) O(Wl) 8700(22) 2186(13) -1359(12) 107(8)* 0 (W2) 7918(36) 964(19) 1177(21) 106(13)* 0(W3) 7027(50) 4484(32) -1136(19) 154(20)* 0 (W4) 9172(47) 4029(27) 92(25) 126(18)* 0(W5) 8939(54) 370(29) -479(27) 93(19)* N (3) 3812(25) 2485(15) 9716(13) 89(9)* C(Il) 4814(57) 3101(33) 10446(26) 235(28)* C (12) 3390(42) 3064(25) 8949(17) 136(17)* C (13) 4470(63) 1387(30) 9355(29) 229(29)* C (14) 2426(51) 2322(39) 10148(25) 232(30)*
*Equivalent isotopic U defined as one third of the trace of the orthogonalised U tensor. 34
Table 15. H-atom coordinates (xlO4) and isotopic thermal parameters (A2XlO3).
Atom X ' Y Z U H (5) 4592 5718 7568 55(11) H (6a) 4935 1831 6839 55(11) H (6b) 6767 2533 6992 55(11) H (6c) 6029 1892 5837 55(11) H (7) 2331 5315 4315 55(11) H (8a) 3311 6677 5540 55(11) H (8b) 2964 6777 6768 55(11) H (9a) 2227 4158 2339 55(11) H (9b) 598 4603 2969 55(11) H (04) 619 6722 5822 55(11) H (51 ) 1298 148 7533 55(11) H (6 ' a) . 807 3874 6667 55(11) H(6'b) -810 3299 7129 55(11) H (6 1C) -875 3401 5893 55(11) H (7 1) 2478 -516 4302 55(11) H(81 a) 1814 -1448 5470 55(11) H(8'b) 2448 -1184 6709 55(11) H (9 1 a) 1962 91 2312 55(11) H (9'b) 3719 -301 2892 55(11) H (04 ') 4540 -161 6160 55(11) H(Ila) 5556 3616 10119 260 H(Ilb) 5541 2578 10813 260 H(Ilc) 4136 3575 10958 260 H(12a) 4433 3214 8597 140 H(12b) 2900 3810 9235 140 H(12c) 2530 2608 8434 . 140 H(13a) 3677 969 8777 230 H(13b) 4569 938 9946 230 H (13c) 5618 1484 9099 230 H(14a) 1437 2225 9607 250 H(14b) 2257 3002 10685 250 H(14c) 2511 1619 10487 250
Table 16. Hydrogen Bonds*. Bond d (A) Bond d (A) O(I1)-H. . .0(4)V 2.82 8() 0(W2)-H...0(3111 2.840( ) 0(1')-H(04)V 1.959( ) O (W4)-H...O (W3)I 2.550( ) O(WS)-H..N(I)III 2.848( ) O(WS)-H. . . O (W2) I 2.567( ) O(Wl)-H. .N d ' ) IV 2.862( ) O(WS)-H.. .O(Wl) I 2.794( ) O (W2) -H. . O (3 ') I 2.868() 0(W4)-H. . .O(Wl) I 2.845( ) 0(W4)-H..0(3)II 2.816( ) O (W4) -H. . .0(W4)VI 2.884 ( ) *Atoms are numbered according to Table 14. The molecules numbering is I: x,y,z; II: l+x,y,z; III: x,y,-l+z; IV: l+x,y,-l+z; V: -x,-y,-z; VI: 2-x,l-y,-z. 35
DISCUSSION
Dunathan1s concept of the favouring of axial over equatorial proton loss and gain due to more efficient a-- tt overlap in the transition state can also be applied to
Co(III) pyridoxylidene glycine. Evidence of distinct positions for glycine protons is clearly seen in the crystal structure of TMAtCo(PGly)2 *• Figure 14 shows the glycine protons in different conformations with respect to the azomethine pi system. H9a protons are in axial orientations largely positioned above the ligands, while
H9b protons remain equatorial in the plane of the ligands. Puckering of the 5-membered metalocycle has forced the alpha carbon out of the ring, tipping one proton into the plane of the.pi system while placing one more orthogonal in orientation. This puckering is identical to that seen in Co(III) 3-methyl salicylidene threonine (Figure 15). Geminal coupling between glycine protons is 20 Hz indicating the angle between protons is 12 approximately of tetrahedral geometry". Movement of carbon, then, is solely responsible for the seperate stereoelectronic environments that control chemical reactivities of the glycine protons. 36
Figure 14. Co(III) pyridoxylidene glycine illustrating axial and equatorial positions of the glycine protons.
Figure 15. Similiar puckering of 5-membered metalocycle of Co(III) 3-methyl salicylidene threonine. 37 Originally, puckering of the 5-membered. metalocycle was believed to be a ramification of a nearby methyl group bonded to the aromatic ring of the neighboring ligand. For various reasons, this steric interaction is unlikely and puckering should be considered wholely electronic in nature. First off, the methyl carbon to alpha carbon distance is in the range of 5-6 angstroms, keeping in mind uncertainties in crystallographic data.
This distance is quite vast for such a steric interaction.
Secondly, as previously stated, the x-ray determined structure of Co(III) salicylidene glycine (no methyl groups on the aromatic rings) shows planarity of the ligands. This is not exactly the case. Two out of the three ligands (one in a special position) show planarity O while the third exhibits a distortion of 14 between the ligand coordination plane and the plane consisting of the alpha carbon, coordinated nitrogen, azomethine carbon, and nearest carbon of the benzene ring (Figure 16). This dihedral angle compares to 24.5* found in the crystal study done on TMA{Co(PGly)2 )• However, ligand distortion in the salicylaldehyde is possibly a construct of crystal packing and likely does not exist in solution. Hydrogen bonds in the crystal form between water molecules and the carboxylic and phenolic oxygens of the anion. This slim 38
Figure 16. Distortion of one ligand of Co(III) salicylidene glycine while the other remains virtually planar.
Figure 17. Distorted A isomer of Co(III) pyridoxylidene valine. 39 evidence suggests perhaps the presence of the methyl group is not absolutely necessary to cause ring puckering.
Lastly and more convincing in arguing that puckering is an electronic manifestation, is the crystal structure done by Capasso, et. al... on Ni(II) pyridoxylidene valine shown in Figure 17. Capasso and associates grew the isomeric A crystals rather than the A. crystals grown in all other instances. Viewing Figure 17, the A isomer places the valine fragment in the most hindered orientation, closest to the methyl group (C6) of the neighboring ligand. Even with the bulky valine group in a position nearest the neighboring methyl, puckering of the
5-membered ring still occurs and in the same direction.
The amount of distortion is small, but since puckering occurs in the direction of the nearby methyl, thereby heightening the steric effect, only electronic motives can explain ring puckering. The natural relief of ring strain as in all 5-membered rings appears just as likely to happen in this situation as well.
Interproton spin-spin coupling across three single bonds and one double bond, J (H-C1=C2-C3-H), where C3 has tetrahedral hybridization, is designated as allylic coupling. There are a number of variables affecting the magnitude of the coupling constant, including 40 substitution at Cl, C2, or C3, the bond order of the
C1-C2 double bond, and angular distortion at either Cl,
C2, or C3, resulting, for example, from incorporation into cyclic structures. But even with as many variables as these, some general correlations between magnitudes of coupling constants and structures of molecules have been made.13'14This is particularly true with the dihedral angle dependence on allylic coupling constants.
Figure 18. Allylic fragment of Co (PGly) demonstrating the dihedral angle dependence.
Na (Co (PGly)2* exhibits an analogous allylic system.
The dihedral angle,#, is measured from the C7-N2-C9 plane as indicated in Figure 18. Affecting the magnitude of the coupling constants is the presence of inorganic cobalt bonded to N2; a shortened N-C double bond in comparison to a normal C-C double bond; and angular distortions of
Cl and C3 due to their inclusion into the 5 and
6-membered metalocycles of the ligand. These variables, very different from allylic systems found in usual
organic molecules, disenable quantitative theoretical 41 formulations previously done, relating size of coupling constants to the dihedral angle dependence. However, in a purely qualitative sense, the magnitude of long range coupling constants between the azomethine proton at C7 and the glycine protons at C9 shows which Of the two glycine protons has the largest dihedral angle; the one tipped more into the plane of the azomethine pi system.
Fundamentally, the alpha carbon proton with the largest dihedral angle will sense the magnetization of the distant allylic proton the greatest, producing the larger splitting. A larger coupling constant, in turn, suggests which proton is more coplanar to the pi system.
Ultimately, a glycine proton with a larger coupling, closer to the pi system, is expected to undergoe deuterium exchange the fastest because of increased tr-ir delocalization. Figure 19 is the nuclear magnetic spectrum of the AB pattern of the glycine protons in D 2O.
The long distant coupling constant of the B proton, the
fast exchanging proton, is I.95Hz compared to a smaller
1.17Hz coupling for the slow A proton. Crystallographic
evidence supports the claim, of non-equivalent proton
environments with respect to the surrounding pi system;
now the previous Karplus relationship, relating proton
environments in solution, agrees with the
crystallographic data. 42
1.17 Hz
1.17 Hz
.95 Hz
I______I______I______I______I______L
Figure 19. AB pattern of the glycine protons along with long distant, allylic coupling by the azomethine proton. Other peaks are the S-CH2 group and a suppressed solvent peak. 43
Crystallographic evidence and NMR allylic coupling constants support different proton orientations, both in crystal form and in solution. In solution, there is no hard evidence, though, that the B protons with larger allylic coupling constants are, in fact, H9a and H.9a1 in the crystal structure. There is the possibility the reverse, H9b and H9b', are closest to the pi system and observe greater long distance splittings. The Kdrplus relationship only states the fast B proton in the NMR spectrum is the most colinear to the pi system.
Activation parameters of exchanging glycine protons determined in the preliminary study by Fischer and Abbott are questionable in accuracy but do reflect general numerical magnitudes found in the current study. Both studies show the fast proton exchanges with a lower AH than the slow proton, but also the fast proton has a considerably reduced entropy of activation. The fast B proton more coplanar to the pi system participates in
greater c t - tt overlap and increased delocalization or resonance stabilization. This overlap assists in bond breakage and lowers the energy necessary to cleave the proton. The entropy of activation for the fast proton has been essentially halved. Still, the relatively large
I negative entropy requires the arrival of base to be colinear to the pi system. This is a sizable restriction 44 on exactly how the proton is abstracted. In enzymes, basic species could be purposely placed in a colinear position, increasing the entropy of activation (less negative) and increasing rates up to a million fold.
What is so rewarding about the Co(III) pyridoxal-glycine Schiff base system are these comparisons made between the two glycine protons.
Usually, comparisons are made between systems analogous in nature but still differing somewhat in manner or quality. For instance, comparisons between the salicylidene (R=H, Figure 5) and 3-methyl salicylidene
(R=CH3 ) systems may be made, but in the end, they are still two different systems. The two glycine protons, when they are on the same structure, however, are nicely comparable one to the other. If we had tried to alter conformations of protons around the alpha carbon, say using some hypothetical steric interaction (size of the steric group), comparison of rates could be made, but, then, perhaps changing steric groups could also contribute inductively. The systems are not really identical. Direct comparisons of the two glycine protons can be made confidently when they are in the same system.
Even more striking in the pyridoxylidene system, but equally inexplicable, is that the fast proton exchanges totally stereospecifically. Formation of a flat carbanion 45 intermediate should exhibit a racemic mixture of
products. If two protons attached to the alpha carbon are
initially present and the fast exchanging proton is lost
to base with carbanion formation, it is possible the
original slow proton would end up in the fast exchanging orientation with deuterium from solvent occupying the
slow position. This inversion of configuration would produce a singlet near the B peak in the NMR spectrum, but this is not observed. Likewise, with retention of
configuration, growth of a singlet peak near the A proton has been observed when the fast proton is exchanged,
leaving behind a deuteron in the same position. For
retention to occur, the rate of exchange must exceed the
rate of racemization.
Finally, Co(III) pyridoxylidene glycine is
justifiably an enzymatic model. One, it kinetically
enhances an amino acid transformation, in this case
simple deuterium exchange, and, two, it completes the
process stereospecifically. Co(III) 3-methyl salicylidene
glycine fulfills the first requirement rather well, the
fast proton exchanges approximately ten times the rate of
the slow proton and infinitely greater than if no
catalytic mechanism were available, but does not compare
to the selectivity of the pyridoxylidene system. In both
salicylidene and pyridoxylidene systems, the methyl group 46 on the aromatic ring controls chemical reactivity around the alpha carbon a great degree. A further important step would be to synthesize the methyl unsubstituted pyridoxal and see precisely what is its effect controlling stereochemistry. This is quite a rigorous synthesis, however. The most beneficial utilitiy of this model system may be its potential for asymetric synthesis.
Already high stereospecificity has been demonstrated,
©penning the possibility of manufacturing optically active amino acids in high yields, especially deuterated amino acids. Only resolution of chirality around the central cobalt atom stands in the way. 47
EXPERIMENTAL
I. Synthesis of NafCo (PGly)
In IOOml of absolute methanol, 0.80 grams NaOH (20 mmol) and 0.75 grams (10 mmol) Glycine were weighed out and dissolved together. Slight heating helped dissolution
of NaOH7 and to this mixture, 2.04 grams (10 mmol) of
Pyridoxal hydrochloride were added and stirred for 30 minutes. To the resulting dark yellow solution, 1.19
grams (5 mmol) CoCl2-OH2O, dissolved in 15ml MeOH, were
added dropwise with vigorous stirring. The solution was
stirred for 15 minutes then placed in an ice bath for
four hours. The product was filtered off as a brown Solid
and vacuum desicated overnight. The procedure yields the
bis(pyridoxylideneglycinato)cobaltate(II) anion in
approximately a 55% yield as its sodium salt.
Oxidation of the Co(II) to the Co(III) complex
occurred as follows. 1.5 grams Co(II) complex were
dissolved in 150ml of absolute methanol with the addition
of 1.5ml 2M NaOH. This was heated to 50°C, and all
remaining solid residues were removed by Buchner 48 filtration. After the addition of 0.75 grams activated charcoal, the solution was aerated for 1.5 hours at room temperature by passing air through a glass frit submerged in a suction flask. The charcoal was removed using a fine frit filter, and the filtrate rotavapped to dryness. This ' left a dark brown product of appropriate composition.
The presence of small amounts of paramagnetic Co (II) in the sample required further purification by extraction. 0.7 grams of the Co(III) complex were dissolved in 7ml of distilled water before addition of
1.5ml 0.SM 8-hydroxyquinoline (Oxine) in chloroform. This process was repeated twice, discarding the previous CHClg extracts on each occasion. The resulting aqueous solution was washed four times with 1.5ml portions of CHCl3 or until the organic phase remained clear. The purified
Co(III) comple was subsequently recovered by rapid evaporation of solvent. Recovery was 80% effective.
Elemental analysis calculated for NaCoC20H20N4Og-
I.SH2O: C/ 43.43; H, 4.16. Found: C, 43.83; H , 4.36. NMR referenced to hexamethyl disiloxane in D2O : 1.548(2—CH3);
4.858(S-CH3); 5.166(glycine H's, average value);
7.641(6-H); 9.017(azomethine C-H). M.W.= 553.13 grams/mole. 1H NMR spectrum is shown in Figure 20. 49
9 8 7 6 5 4 3 2 1 0 ppm Figure 20. H NMR spectrum of Co(III) pyridoxylidene glycine.
2. Preparation of COg/DCO^ buffers.
Starting solutions of I.OM KCl and 0.1M Na2C02 in carbonate were made using >98.5% D^O. 1.0ml of 0.1M Na2C03 was acidified with DCl to the appropriate pH and diluted to 2ml with D2O. 0.5ml of buffer was then added to 2ml of the I.0M KCl, leading to the ionic strength adjusted (0.8M) buffer. Total carbonate concentration, COg + DCOg , equaled 0.01M. Carbonate concentrations in solution were calculated from the following two equations: 50
[CO3=] + [DCO3-I = [CO3=It =0.01M 5)
[D+ ] [CO3=I 6) K = a [DCO3-]
Substituting:
K (0.01) - K [CO =] = [D+ H C O 0 = ] I) a a 3
16 Replacing temperature dependent equilibrium constants and experimentally determined deuteron concentrations in equation 7), carbonate concentrations in solution are calculated.
3. Collection of kinetic data.
To monitor the disappearance of the glycine protons, spectrophotometric data were collected using nuclear magnetic resonance (NMR) on a Bruker WM 250. I.3mg
Co(III) complex dissolved in 0.60ml buffer solution was the standard concentration used for the kinetic runs.
Buffers mixed at a particular pH fell approximately
0.4-0.5 pH units after addition of the potassium chloride and after addition of the Co(III) complex due to dissociation of proton from the pyridinium nitrogen of the complex. Buffers remained constant in pH during 51 kinetic runs bid did fall initially when reactants were mixed. Integration of peak heights provided raw kinetic data/ keeping the baseline and maximum integration height as flat as possible. Error in integration is approximately 10%, but error in integration of the fast B proton is greater than 10% due to the inability to correctly judge the inflection point between A and B peaks. Error in pseudo firtst order rate constants for the fast proton is approximately 15%. Kinetic runs were typically carried out to 50-70% completion. The pD values of buffer solutions were determined with a glass electrode on a Brinkmann E512 from pD=pH+0.4, where pH is the observed pH of the solution on completion of the
kinetic run. The S-CH2 group of the methoxy group on the pyridine ring was used as the internal reference with no
change in peak area during the kinetic runs.
4. Single-crystal X-ray diffraction analysis of
TMA{Co(PGly)2>*3H20
Suitable crystals for diffraction were grown in HgO
by slow diffusion of acetone. A dark brown crystal
fragment was chosen, measuring approximately
0.1x0.2x0.6mm, and mounted atop a glass fiber. Data
collection was carried out on a Nicolet R3mE automated 52 diffractometer using 0/20 scans and graphite monochromated MoKa radiation (A=O.71069A). Unit cell constants were determined to be a=8.514(3)A, b=12.755(4)A, c=14.0 89 (7) A, a =98.88 (3) °, /3 = 94.66 (3)% 7
=91.17(3)"for a cell of triclinic symmetry by the method of Campana. The space group was identified as PI, and for C 4H12ncQ (c IqH i o ^ O 4) 2’3H20, a cell mass of 1263.1 gave a calculated density of 1.39g/cm. Check reflections, showed no reduction in intensity over the course of the data collection.
The structure was solved through a Patterson synthesis that revealed the location of the Co atom and subsequent least squares refinements to determine positions of all other non-hydrogen atoms. The structure was refined to a final R value of 0.117, assigning anisotropic thermal values to the cobalt, coordinating oxygens and nitrogens, water oxygens, and atoms of the tetramethylammonium ion. X-ray determinations of molecules of near identical structure and composition exhibit R values of 0.10, 0.05, 0.16, 0.16, 0.14, and
O. ^ ? -10'18 Hydrogen atoms were placed in idealized positions with isotopic thermal parameters. All structural determinations and refinement calculations were carried 19 out with the SHELXTL package by Nicolet. The necessary
X-ray data are summarized below. 53
Table 17. Crystallographic data.
Space group Pl T = 298K a = 8.514(3) R = 0.117 b = 12.755(4) Rw = 0.118 c = 14.089(7) F(OOO) = 657.89
a = 98.88(3) Scan type = w
/3 = 94.66 (3) Scan speed = 6.0-29.3 /min
7 = 91.17(3) 20 range = 3-40
V = 1505.83 Scan range [2.0+ I
Z = 2 Reflections collected = 3007
D (calc) = 1.39 Unique reflections = 2751 X X(MoKa) = 0.71069A Observed reflections = 1769 — 3. = I . OOcm # of L.S. parameters = 271.
Reflections measured = +9,+13,±14 54
REFERENCES
1. Gyorgy, P. Nature. 1934,133, 498.
2. Metzler, D.E.;E.E. Snell. J. Amer. Chem. Soc. . 1952,74, 979.
3. Metzler, D.E.;M. Ikawa;E.E. Snell. J. Amer. Chem. Soc. 1954,76, 648.
4. Dunathan, H.C. Proc. Natl. Acad. Sci., USA. 1966,55, 712.
5. Corey, E.J.?R.A. Sneen. J. Amer. Chem. Soc. 1956,78, 6269.
6. Malhotra, S.K.;H.J. Ringold. J. Amer. Chem. Soc. 1964,86, 1997.
7. Yu N. Belokon; et. al. Tetrahedron. 1977,33, 2551.
8. Aleksandrov,G.G.;Y.T. Struchkov;Y.N. Belokon. Zh. Struktur Khimii. 1975,16, 875.
9. Fischer, J.R.;E.H. Abbott. J. Amer. Chem. Soc. 1979,101, 2781.
10. Capasso, S.;F. Giordano;C. Mattia;L. Mazzarella; A. Ripamonti. J. Chem. Soc. D.T. 1974, 2228.
11. Willstadter, E.;T.A. Hamor ?J.L. Hoard. J. Ame Chem. Soc. 1963,85, 1205.
12. Pavia, D.L.;G.M. Lampman;G.S. Kriz."Introduction to Spectroscopy";W.B. Saunders Company: Philadelphia, 1979; p 120. NMR geminal coupling constant done on a ■ Bruker WM 250 with I.3mg sample in 0.6ml D20.
13. Karplus, M. J. Chem. Phys. 1959,30, 11.
14. Karplus, M. J. Amer. Chem. Soc. 1963,85, 2870.
15. Barfield, M.;R.J. Spear;S. Sternhell. Chemical Reviews. 1976,76, 593.
16. Paabo, M.;R.G. Bates. Analytical Chemistry. 1969,41, 283. 55
17. Campana, C.F.;D.F. Shepard;W.M. Litchman. Inorganic Chemistry. 1981,20, 4039.
18. Cutfield, J.F.;D. Hall;T.N. Waters. Chemical Communications. 1967,14, 785.
19. Programs used for data reduction, Fourier synthesis, direct methods, Patterson, least-squares planes calculations, and calculations of hydrogen positions are those described in "Nicolet SHELXTL Structure Determination Manual", Sheldrick, G.M., Ed., Nicolet XRD Corp., Fremont, CA, 1980. MONTANA STATE UNIVERSITY LIBRARIES 111I111III illIII
CO 7152 100 1560( 7
?%.in N378 Sy44 Sykes , A. G. cop.2 Molecular structure and reactivity of Vitamin ...
ISSUED TO _ DATE
---- MAITT N378 Sy 44 cop. 2